meme-generation
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses shell commands to execute its internal Python script for meme generation. These commands take user-supplied captions as arguments, which are then processed by the script.
- [EXTERNAL_DOWNLOADS]: The skill fetches meme templates and metadata from the imgflip API and content delivery network. These are well-known services for meme content, and the downloads are limited to image files and JSON metadata for legitimate functionality.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text to generate memes and can ingest image descriptions. While this provides a surface for indirect prompt injection, the impact is low as the data is used primarily for text rendering and visual analysis by the agent.
- Ingestion points: User-supplied captions and topics in
SKILL.md. - Boundary markers: None explicitly defined for the caption arguments.
- Capability inventory: File system writes (saving images) and subprocess execution of the internal Python script.
- Sanitization: The script performs basic text wrapping and scaling but does not include advanced sanitization of the input strings.
Audit Metadata